EP2507884B1 - Inverter for high voltages - Google Patents

Description

The invention relates to a submodule for forming an inverter with a first subunit, which has a first energy store, a first series circuit of two power semiconductor switching units connected in parallel to the first energy store, which each have a power semiconductor which can be switched on and off in the same forward direction and are conductive in each case counter to said forward direction, and a first terminal, which is connected to the potential point between the power semiconductor switching units of the first series circuit, and a second subunit, the second energy storage, a second energy storage parallel-connected second series connection of two power semiconductor switching units, each having a turn on and off power semiconductor in the same forward direction and in each case are conductive opposite to said forward direction, and having a second terminal which is connected to the potential point t is connected between the power semiconductor switching units of the second series circuit.

The invention further relates to a converter for example for high voltage applications with power semiconductor valves, each extending between an AC voltage terminal and a DC voltage terminal and forming a bridge circuit, each power semiconductor valve having a series connection of bipolar submodules and each submodule has at least one energy store and at least one power semiconductor circuit.

The aforementioned submodule is from the EP 1 497 911 A2 already known. There, a direct converter is described with the converter branches, which consist of a series connection of identical bipolar submodules. The submodules have two subunits, each having a unipolar capacitor and a series circuit of two IGBTs, each IGBT a freewheeling diode is connected in parallel in opposite directions. A first submodule connection terminal is connected to the potential point between the IGBTs of the first subunit and the second submodule connection terminal is connected to the potential point between the IGBTs of the second subunit. The two capacitors of the subunits are coupled to one another via connecting means, wherein the connecting means have at least two IGBTs, each with a parallel freewheeling diode.

In the DE 101 03 031 A1 An inverter is described which has power semiconductor valves connected together in a bridge circuit. Each of these power semiconductor valves has an AC voltage terminal for connecting a phase of an AC voltage network and a DC voltage terminal, which can be connected to a pole of a DC intermediate circuit. In this case, each power semiconductor valve consists of a series circuit of two-pole submodules, each having a unipolar storage capacitor and a power semiconductor circuit in parallel with the storage capacitor. The power semiconductor circuit consists of a series circuit in the same direction oriented switched on and off power semiconductor switches, such as IGBTs or GTOs, each of which a freewheeling diode is connected in parallel in opposite directions. One of two terminals of each submodule is connected to the storage capacitor and the other terminal to the potential point between the two can be turned on and off Power semiconductor switches connected. Depending on the switching state of the two controllable power semiconductors, either the capacitor voltage dropping across the storage capacitor or else a zero voltage can be applied to the two output terminals of the submodule. Due to the series connection of the submodules within the power semiconductor valve, a so-called DC voltage impressing multi-stage converter is provided, wherein the height of the voltage stages is determined by the height of the respective capacitor voltage. Multi-stage or multi-point converters have the advantage over the two- or three-stage converters with central capacitor banks that high discharge currents are avoided in the event of a short circuit on the DC side of the converter. In addition, the complexity of filtering harmonics over two- or three-phase converters is reduced in the case of multistage converters.

Multipoint converters are also preferred for the construction of spatially extended branched DC voltage networks, which are required in particular in so-called offshore wind farms and in connection with solar power networks in desert areas.

However, an important prerequisite for the use of inverters in these areas is a secure control of short circuits in the DC voltage network. Cheap mechanical switches for extremely high DC voltages, which can switch high fault currents under load, are not available due to fundamental physical problems. The technically achievable switch-off times and the switching overvoltage of mechanical switches are also disturbing.

The EP 0 867 998 B1 describes the use of electronic power semiconductor switches in the DC link of a high voltage DC power transmission system. The use of power semiconductor switches at DC voltages of a few hundred kilovolts, however, has the disadvantage that the high voltage makes a large number of power semiconductors connected in series necessary. But this also introduces a high passage loss on these components. In addition, overvoltage limiters must be provided in parallel to the power semiconductors, whereby the effort is additionally increased. The surge limiters usually have no ideal Begrenzerkennlinien so that the number of series-connected power semiconductors must be designed even higher than the nominal voltage would actually require. As a result of this oversizing, the forward losses increase even further.

The WO 2008/067786 A1 describes a multi-stage converter with series connections of submodules, each submodule having a thyristor next to a capacitor in parallel to a power semiconductor circuit. The thyristor is connected in parallel with a free-wheeling diode of the power semiconductor circuit, which leads to the entire short-circuit current in the event of a fault. In the event of a short circuit, the parallel thyristor is ignited so that the freewheeling diode is relieved.

In addition to the above-mentioned applications in the field of electric power transmission and distribution, DC multipoint converters are, of course, also excellently suited for use in the field of drive technology.

The above-mentioned multi-point or multi-stage converters have the disadvantage that a short-circuit current across the inverter can not be limited without additional measures in both directions, so that the semiconductors of the inverter and external components are endangered or destroyed in the short circuit.

The object of the invention is to provide a submodule and a converter of the type mentioned above, with which occurring in the event of a fault occurring short-circuit currents effectively limited and damage to the system can be safely avoided and which is also inexpensive. In addition, faulty sections of a DC voltage network should be de-energized as quickly as possible and separated in this way from the rest of the DC network.

Finally, in the event of a short circuit on the DC side of the converter, the currents on its AC side should be influenced as little as possible, and tripping of the AC side mechanical switches should be avoided.

Starting from the submodule mentioned above, the invention achieves this object by virtue of the first subunit and the second subunit being connected by connecting means which connect an emitter connection branch which connects an emitter of a first power semiconductor switching unit of the first series circuit to an emitter of a first power semiconductor switching unit of the second series circuit and in which a potential separation diode is arranged, a collector connection branch connecting a collector of the second power semiconductor switching unit of the first series circuit to a collector of the second power semiconductor switching unit of the second series circuit and in which a potential separation diode is arranged, and a Having a switching branch, in which a switching unit is arranged and connects the cathode of the potential separation diode of the emitter connection branch with the anode of the potential separation diode of the collector connection branch.

Based on the aforementioned converter, the invention solves this problem in that the submodule is a submodule according to the invention.

According to the invention, two subunits each having an energy store, for example a capacitor, and a series connection of two power semiconductor switching units are connected to one another via connecting means. Deviating from the prior art, the connection means are designed such that, with suitable control of the power semiconductor switching units, a current flow between the two connection terminals of the submodule according to the invention must always occur via at least one energy store. The affected energy storage device always builds up a countervoltage independent of the polarization of the clamping current, which quickly reduces the current flow. The selected switching state according to the invention depends on the topology of the connection means and their components.

According to the invention, a high short-circuit current can be controlled without external additional switches. In contrast to the prior art, it is ensured in the context of the invention that high short-circuit currents can be avoided quickly, reliably and effectively by the converter itself in both directions. Additional switches, for example in the DC voltage circuit which is connected to the converter, or else semiconductor switches connected in parallel with a power semiconductor of the submodule, have become superfluous in the context of the invention. In case of error take almost exclusively the submodules according to the invention on the released energy, so that it is completely absorbed. The energy intake has a counter tension in the wake and can be dimensioned in a defined and desired manner, so that unfavorably high voltages are avoided. In addition, according to the invention no energy storage must be charged controlled to restart the inverter. Rather, the inverter according to the invention can resume normal operation at any time.

According to the invention, the connection means comprise an emitter connection branch which connects an emitter of a first power semiconductor switching unit of the first series circuit to the emitter of a first power semiconductor switch unit of the second series circuit and in which a potential isolation diode is arranged. Further, a collector connection branch is provided, which connects a collector of the second power semiconductor switching unit of the first series circuit with a collector of the second power semiconductor switching unit of the second series circuit and in which also a potential separation diode is arranged. The connection means further comprise a switching branch in which a switching unit is arranged and which connects the cathode of the potential separation diode of the emitter connection branch to the anode of the potential separation diode of the collector connection branch. The emitter of a power semiconductor switching unit is also referred to as source or cathode.

The connecting means expediently have a switching unit. This switching unit is in said selected state, for example in its interruption position. Deviating from this, however, it is also possible according to the invention that the switching unit in the selected switching state in their Passage position is. The design of the switching unit is fundamentally arbitrary within the scope of this further development of the invention. For example, this may be a mechanical switching unit, a suitable semiconductor switch, or a power semiconductor switching unit that is similar to the other power semiconductor units of the converter. The design of the power semiconductor switching units will be discussed in more detail later.

Conveniently, the connection means comprise at least one potential isolation diode arranged to maintain a voltage difference between the first subunit and the second subunit. According to this advantageous development, it is possible to increase the number of achievable voltage levels. For example, it is possible to generate the sum of the voltages dropping across the first energy store and the voltages dropping across the second energy store at the connection terminals of the submodule. In addition, in this embodiment of the invention, the possibility depending on the switching state of the power semiconductor switching units only one, so either to generate the voltage dropping at the first energy storage or the second energy storage at terminals. In this way, the first and the second subunit can be treated in terms of control technology like two submodules according to the prior art. Thus far established control methods can also be applied to the submodule according to the invention.

Moreover, it is advantageous that the connecting means have at least one damping resistor. The damping resistor (s) support the energy storage devices to absorb energy in the event of a fault. For this purpose, the damping resistors are so with the remaining components of the connecting means interconnects that in the said selected switching state, a current flow, at least partially, also via the damping resistances, regardless of the polarity of the clamping current.

According to a related expedient further development, in each case a damping resistor is arranged in the emitter connection branch and in the collector connection branch. As already stated, the switching unit of the switching branch is basically arbitrary. It is essential that the switching unit between a breaker position in which it interrupts a current flow, and a passage position in which it is conductive, can be switched back and forth. It is thus possible, for example, to use a mechanical power switch, a cost-effective semiconductor switch or a power semiconductor switching unit as the switching unit, which is similar to the other power semiconductor switching units of the submodule. Other controllable power semiconductors can be used as a switching unit in the context of the invention.

As already stated, the selected switching state is achieved according to this expedient further development when all the power semiconductor switching units and the switching unit are in their breaker position. Of the Clamping current is now performed in any case via at least one energy storage or a damping resistor.

In any case, the switching unit should be selected so that the power loss resulting from it during normal operation of the submodule is as low as possible.

If all the power semiconductor switching units of the submodule are designed identically, in other words all the semiconductor switches are identical, they have a uniform blocking voltage and structure. This is advantageous at high voltages because only a few semiconductor switches are suitable for extremely high voltages and powers. Uniform submodule placement makes it possible to use the most suitable and powerful semiconductors.

Expediently, each power semiconductor switching unit has a power semiconductor that can be switched on and off, to which a freewheeling diode is connected in parallel in opposite directions. Such turn-off power semiconductors are, for example, marketable IGBTs or GTOs and the like. These power semiconductors are usually used with freewheeling diode connected in parallel in opposite directions. However, according to the invention, reverse-conducting power semiconductors can also be used. Separate freewheeling diodes are then unnecessary.

Expediently, each energy store is designed as a capacitor and in particular as a unipolar storage capacitor.

Figur 1

ein Ausführungsbeispiel des erfindungsgemäßen Umrichters schematisch verdeutlicht und

Figur 2

ein Ausführungsbeispiel der erfindungsgemäßen Submodule genauer dargestellt.

Further advantages and embodiments are the subject of the following description of embodiments with reference to enclosed figures of the drawing, wherein the same References to like components and reference

FIG. 1

an embodiment of the inverter according to the invention schematically illustrated and

FIG. 2

an embodiment of the submodules according to the invention shown in more detail.

FIG. 1 shows an embodiment of the inverter 1 according to the invention in a schematic representation. It can be seen that the converter 1 has power semiconductor valves 2 which are connected to one another in a bridge circuit. Each of the power semiconductor valves 2 extends between an AC voltage terminal L ₁ , L ₂ , L ₃ and a DC voltage terminal 3 ₁ , 3 ₂ , 3 ₃ and 4 ₁ , 4 ₂ , 4 _third The DC voltage terminals 3 ₁ , 3 ₂ , 3 ₃ are connected via a positive pole terminal 5 with a positive pole and a negative pole terminal 6 with a negative pole of a DC voltage network not shown figuratively.

The AC voltage terminals L ₁ , L ₂ and L ₃ are each connected to a secondary winding of a transformer, whose primary winding is connected to a likewise not shown figuratively AC mains. For each phase of the alternating voltage network, an AC voltage terminal L ₁ , L ₂ , L _{3 is} provided. In the embodiment shown, the AC voltage network is three-phase. Thus, the inverter 1 has three AC voltage terminals L ₁ , L ₂ and L ₃ . Between the AC voltage terminal L ₁ , L ₂ , L ₃ and the transformer mechanical circuit breakers are expediently provided to disconnect the AC power supply from the inverter 1 in case of failure. The circuit breakers are in FIG. 1 also not shown.

In the exemplary embodiment shown, the converter 1 is part of a high-voltage direct-current transmission system and is used to connect AC voltage networks in order to transmit high electrical power between them. It should be noted, however, that the converter can also be part of a so-called FACTS system, which serves to stabilize the network or to ensure a desired voltage quality. In addition, use of the inverter is also according to FIG. 1 and 2 possible in drive technology.

In FIG. 1 Furthermore, it can be seen that each power semiconductor valve 2 has a series connection of submodules 7 and a throttle 8. Each submodule 7 has two connection terminals x1 and x2.

FIG. 2 shows an embodiment of the submodule 7 according to the invention in more detail. It should be noted at this point that all in FIG. 1 schematically shown submodules 7 are constructed identically. FIG. 2 shows therefore the structure of all submodules 7 of the inverter 1 representative of a submodule. 7

The submodule 7 according to FIG. 2 has a first subunit 9 and a second subunit 10, which are framed by a dashed line and constructed identically. Thus, the first subunit 9 comprises a first series circuit 11 of power semiconductor switching units 12 and 13, which in the embodiment shown each have an IGBT 14 or 15 as turn on and off power semiconductors and one freewheeling diode 16 and 17, the respectively associated IGBT 14, 15th in opposite directions is connected in parallel. The IGBTs 14, 15 have the same forward direction, so are oriented in the same direction. The potential point between the power semiconductor switching units 12 and 13 is connected to a first terminal x2. The series circuit 11 is connected in parallel to a first capacitor 18 as a first energy store, at which the voltage U _C1 drops.

The second subunit 10 includes a second series circuit 19 of a first power semiconductor switching unit 20 and a second power semiconductor switching unit 21, each having an IGBT 22 and 23 as turn on and off power semiconductors. The IGBTs 22, 23 have the same forward direction in the series circuit 19, so that the power semiconductor switching units 20 and 21 are oriented in the same direction. Each IGBT 22 or 23 of the second series circuit 19 is a freewheeling diode 24 and 25 connected in opposite directions in parallel. The second series circuit 19 is connected in parallel to a second capacitor 26, at which the voltage U _C2 drops. The potential point between the power semiconductor switching units 20 and 21 is connected to the second terminal x1.

The subunits 9 and 10 are linked via connecting means 27. The connection means 2 have an emitter connection branch 28 and a collector connection branch 29. The emitter connection branch 28 connects the emitter of the IGBT 15 of the first series circuit 11 to the emitter of the IGBT 23 of the second series circuit 19. The collector connection branch 29, on the other hand, connects the collector of the IGBT 14 of the first series circuit 11 to the collector of the IGBT 22 of the second one Series connection 19. In the emitter connection branch 28, a potential separation diode 30 and a limiting resistor 31 are arranged. The collector connection branch 29 also has a potential separation diode 32 and a limiting resistor 33. The emitter connection branch 28 is connected to the collector connection branch 29 via a switching branch 34, in which a switching unit 35 is arranged. In the embodiment shown, the switching unit is realized as a power semiconductor switching unit 35 and includes an IGBT 36 and a freewheeling diode 37 connected in parallel thereto. In this case, the switching branch 34 connects the cathode side of the potential separation diode 30 with the anode side of the potential separation diode 32, which between the said anode and the Switching branch 34 arranged limiting resistor 33 has been neglected.

The operation of the circuit of the submodules 7 will be explained below. First, it should be noted that the required reverse voltage of all power semiconductors, so both the freewheeling diodes 16, 17, 24 and 25 and the on and off power semiconductor switches 14, 15, 23 and 23, after the maximum voltage of the two unipolar storage capacitors 18 and 26, which is the same in the chosen embodiment. In this way, a disadvantageous oversizing of the blocking voltages of said power semiconductors is avoided.

Overall, it is possible to differentiate between a plurality of switching states which differ from one another with regard to the clamping voltages U _x .

In an exemplarily selected switching state 1, the terminal voltage _x falling at the connection terminals x2 and x1 is equal to zero irrespective of the direction of the clamping current. In this switching state, the IGBTs 15, 22 and 36 are in their forward position, in which a flow of current in the forward direction through the respective IGBT is possible. The remaining IGBTs, ie the IGBTs 14 and 23, are located however, in their blocking position, so that a current flow through these IGBTs is interrupted. With positive current direction i _x (i _x positive), the in FIG. 2 is indicated at the first terminal x2 by the arrow, the power semiconductors 15, 37 and 22 are energized. With negative current direction (i _x negative), the power semiconductors 24, 36 and 17 are live.

The following table summarizes the preferred switching states used. switching status i _x IGBT 15 IGBT 14 IGBT 23 IGBT 22 IGBT 36 U _X W _CI W _C2 1 negative 1 0 0 1 1 0 0 0 positive 1 0 0 1 1 0 0 0 2 negative 0 1 0 1 1 + U _C1 -1 0 positive 0 1 0 1 1 + U _C1 +1 0 3 negative 0 1 1 0 1 + (U _C1 + U _C2 ) -1 -1 positive 0 1 1 0 1 + (U _C1 + U _C2 ) +1 +1 4 negative 1 0 1 0 1 + U _C2 0 -1 positive 1 0 1 0 1 + U _C2 0 +1 5 negative 0 0 0 0 0 - (U _C1 + U _C2 ) / 2 +1 +1 positive 0 0 0 0 0 + (U _C1 + U _C2 ) +1 +1

The columns W _C1 and W _C2 are intended to clarify whether the storage capacitors 18 and 26 absorb or deliver energy, where +1 stands for the intake and -1 for the delivery of energy.

It can be seen from the table that in the switching states 2, 3 and 4 a positive voltage is always generated at the terminals x2 and x1. This applies regardless of the direction of the clamping current. For example, the capacitor voltage U _C1 or the capacitor voltage U _C2 or else the sum of the capacitor voltage U _C1 + U _C2 at the connection terminals can drop.

wobei U_R dem Spannungsabfall an den Dämpfungswiderständen 32 und 30 entspricht.In the switching state 5, all controllable power semiconductors, ie the IGBTs 14, 15, 22, 23 and 36, are in their breaker position, so that a current flow through the IGBTs is interrupted. In this switching state, the terminal voltage U _x always forms a countervoltage independent of the polarity of the clamping current i _x , so that the submodule 7 always picks up energy. In the negative current direction, i _x negative, a negative reverse voltage is generated by the parallel connection of the storage capacitors 26 and 18 and by the voltage drop across the damping resistors 30 and 32. If the capacitor voltages U _C1 and U _C2 do not match exactly, they are automatically balanced. In switching state 5 applies to a good approximation $$U_x=\frac{-\left(U_{C1}+U_{C2}\right)}{2}-U_R$$

where U _R corresponds to the voltage drop across the damping resistors 32 and 30.

erzeugt. Auch hier kann ein Stromfluss nur unter Aufladung der Speicherkondensatoren 18 beziehungsweise 25 erfolgen. Dabei ist es vorteilhaft, dass der auftretende Strom über beide Kondensatoren geführt wird, da an diesen dann eine geringere Überspannung auftritt, als wenn nur ein Kondensator die Energie aufnehmen müsste.A positive current direction becomes a positive reverse voltage $$U_x=+\left(U_{C1}+U_{C2}\right)$$

generated. Here, too, a current flow can take place only when the storage capacitors 18 and 25 are charged. It is advantageous that the occurring current is passed through both capacitors, as then a lower overvoltage occurs at this, as if only one capacitor would have to absorb the energy.

It can also be seen from the above table that with the submodule 7 and its two subunits 9 and 10, the same output voltages can be generated at the output terminals as in two series-connected submodules according to the prior art ( DE 101 03 031 A1 ). The subunits 9, 10 virtually each correspond to a submodule according to the prior art. In other words, the inventive submodule according to FIG. 2 in the same manner as two submodules according to the prior art. All known control methods are therefore still applicable. In the narrower sense, however, the secondary condition exists that the number of submodules connected in series according to the prior art always has to be an even number. However, in high voltage applications, the number of submodules connected in series is so large that this constraint is insignificant.

The switching state 5 can be used in case of failure for complete power reduction. If all submodules 7 are converted into this switching state, the branch currents of the converter 1 and, as a result, the alternating voltage and DC side currents as a result of the sum of the reverse voltages of all series-connected submodules 7 are reduced very rapidly to zero. The speed of this current reduction results from the above-mentioned counter-voltage and the inductances summarily present in the circuits. In the embodiment shown, it is typically of the order of a few milliseconds.

The dead time until the beginning of the current reduction depends essentially on the response time of the switching unit 35. If a power semiconductor switching unit according to FIG. 2 used, this dead time is negligible. The dead time is then essentially the inertia of the owing to various measuring sensors and current transformers, with the aid of which a fault is detected. This inertia of these measurements is currently typically in the range of a few tens of microseconds.

The advantages of the submodule according to the invention and of the converter 1 according to the invention can be summarized as follows: On the one hand, the time span until the complete removal of a short-circuit current occurring in the event of a fault is very short. Thus, provided on the AC side of the inverter 1 switches must not be triggered only. Both the AC-side and the DC-side current exceed the rated current only insignificantly. The power semiconductors of the submodules need not be protected with thyristors or other bridging elements as in the prior art. The reliability of the power cutoff is very high because redundancy is ensured in the power semiconductor valves of the inverter 1 by the large number of submodules connected in series. In connection with the reliability is still stated that the inverter 1 is constantly in operation with all its components and metrologically monitored continuously. Such reliability is not given in comparable devices for power dissipation in case of failure, which are activated only in such an error case.

Another significant advantage of the invention is that at any time a "switch back" in normal operation is possible, so that even if incorrect unnecessary triggering or detection, the negative effects on the system operation can be minimized.

With the aid of an inverter 1 according to the invention, it is also possible, even in a branched DC voltage network to quickly bring the DC currents to zero. In this way, in the DC voltage circuit, an electroless separation, for example with vacuum interrupters or anti-parallel thyristors possible. In branched DC networks, of course, the other converters that are connected to the DC power supply must reduce the power, so quickly go into the switching state 5 of the submodules 7. A faulty network section of the DC voltage network can thus be disconnected from the remaining DC voltage network in a simple and cost-effective manner by means of known mechanical switches. The faulty network section can now be "paused" for the purpose of dionization or fault localization and later driven up by its assigned converter. In a very short time, the remaining inverters can put the entire direct current network back into operation.

Claims (10)

1. Sub-module (7) for forming a converter (1) comprising

   - a first sub-unit (9), which has

   - a first energy store (18),

   - a first series circuit (11) - connected in parallel with the first energy store (18) - formed by two power semiconductor switching units (12, 13), which each have a power semiconductor (14, 15) that can be turned on and off and have the same forward direction, and the power semiconductor switching units (12, 13) are each conductive counter to said forward direction, and

   - a first connection terminal (x2), which is connected to the potential point between the power semiconductor switching units (12, 13) of the first series circuit (11), and a second sub-unit (10), which has

   - a second energy store (26),

   - a second series circuit (19) - connected in parallel with the second energy store (26) - formed by two power semiconductor switching units (20, 21), which each have a power semiconductor (22, 23) that can be turned on and off and have the same forward direction, and the power semiconductor switching units (20, 21) are each conductive counter to said forward direction, and

   - a second connection terminal (x1), which is connected to the potential point between the power semiconductor switching units (20, 21) of the second series circuit (19),

   characterized in that
   the first sub-unit (9) and the second sub-unit (10) are connected to one another via connecting means (27) have an emitter connecting branch (28), which connects an emitter of a first power semiconductor switching unit (13) of the first series circuit (11) to an emitter of a first power semiconductor switching unit (21) of the second series circuit (19) and in which a potential isolating diode (30) is arranged, a collector connecting branch (29), which connects a collector of the second power semiconductor switching unit (12) of the first series circuit (11) to a collector of the second power semiconductor switching unit (20) of the second series circuit (19) and in which a potential isolating diode (32) is arranged, and a switching branch (34), in which a switching unit (35) is arranged and which connects the cathode of the potential isolating diode (30) of the emitter connecting branch (28) to the anode of the potential isolating diode (32) of the collector connecting branch (29).
2. Sub-module (7) according to Claim 1,
   characterized in that
   the switching unit (35) is a mechanical switching unit, a semiconductor switch or a power semiconductor switching unit.
3. Sub-module (7) according to any of the preceding claims,
   characterized in that
   the connecting means (27) have at least one damping resistor (31, 33).
4. Sub-module (7) according to Claim 3,
   characterized in that
   damping resistors (31, 33) are arranged in the emitter connecting branch (28) and in the collector connecting branch (29).
5. Sub-module (7) according to any of the preceding claims,
   characterized in that
   in a selected switching state all the power semiconductor switching units (12, 13, 20, 21, 35) are in their interrupter position, wherein a current flow takes place between the first connection terminal (x2) and the second connection terminal (x1) in both directions only via the first energy store (18) and/or the second energy store (26).
6. Sub-module (7) according to Claim 5,
   characterized in that
   the switching unit (35) is in its interrupter position in the selected switching state.
7. Sub-module (7) according to any of the preceding claims,
   characterized in that
   the power semiconductor switching units (12, 13, 20, 21, 35) are reverse conducting power semiconductor switches that can be turned on and off.
8. Sub-module (7) according to any of Claims 1 to 7,
   characterized in that
   each power semiconductor switching unit (12, 13, 20, 21, 35) has a respective power semiconductor (14, 15, 22, 23, 26) that can be turned on and off, with which power semiconductor a freewheeling diode (16, 17, 24, 25, 37) is connected in parallel in the opposite sense.
9. Sub-module (7) according to any of the preceding claims,
   characterized in that
   each energy store is a unipolar storage capacitor (18, 26).
10. Converter (1) particularly for high-voltage applications comprising power semiconductor valves (2), which in each case extend between an AC voltage connection (L₁, L₂, L₃) and a DC voltage connection (3₁, 3₂, 3₃, 4₁, 4₂, 4₃) and form a bridge circuit, wherein each power semiconductor valve (2) has a series circuit formed by two-pole sub-modules (7) and each sub-module (7) has at least one energy store (18, 26) and at least one power semiconductor circuit (11, 19),
    characterized in that
    the sub-module is a sub-module (7) according to any of the preceding claims.

Images